Science: What sends a lemming over the edge?: Malcolm Smith reports on a likely explanation for the mass suicides of these northern rodents

Devotees of the computer game 'Lemmings' know how difficult it is to ensure that these stubby-tailed rodents arrive safely at their destination. In the wild - in Scandinavia and north-west Russia - it is impossible.

These little rodents, no more than five inches in length, undergo one of the best-known cyclical fluctuations in their numbers that nature ever invented. Wildlife documentary after wildlife documentary has shown close-up footage of thousands of the little, vole-like mammals careering headlong into rivers and the sea - frequently to certain death. The question is, why?

Lemmings are cold-climate mammals. They live in moist, stony, tundra vegetation consisting of sedges, willow scrub and dwarf birch. For half the year - often longer - their environment is frozen and snow-covered.

Northern Europe and north-west Asia have three species, of which the best-known is the Norway lemming - the film star. Usually nocturnal, it has a stumpy tail, short legs, and a round body covered in thick, yellow-brown fur patterned with dark brown streaks and patches.

Every three to five years, lemming numbers peak, then suddenly decline again. At these population peaks, the excess number of rodents often move down from their mountain tundra homes to valleys and spread out, sometimes en masse. Eventually, some reach the sea and attempt to continue this frenzied dispersal. With uncanny timing, the wildlife film-makers are usually on the spot.

All sorts of theories to explain these enormous population fluctuations have been put forward. They include predation, nutrition, disease, parasites, competition and genetic changes. None has been proven. Another theory - that lemming population swings are related to chemical changes in the plants they eat - had not been tested - until recently.

Tarald Seldal and Goran Hogstedt of the University of Bergen, and Knut-Jan Andersen of Haukeland Hospital in Bergen, have done just that. Their findings provide very strong circumstantial evidence that chemicals produced by the plants grazed by Norway lemmings control the ups and downs these rodents have to cope with.

Lemmings graze plants like stiff sedge and cotton grass. Both are abundant in the Scandinavian tundra, and both produce defensive chemicals when damaged. The heavier the grazing - in other words the damage - the more of these chemicals they produce. Tundra plants are not unique in this; many plants produce such defensive substances.

Some of these chemicals are proteins which inhibit the activity of proteases (enzymes which break down protein in food) in the intestines of mammals. The most important feature of these inhibitors is that they put a stop to the action of trypsin, a protease secreted by the pancreas.

Grazing mammals ingesting these inhibitors can't digest proteins in their food. To make matters worse, reduced protein digestion stimulates more production - by the pancreas - of proteins which, in turn, are inhibited and lost in the faeces. These enzymes contain large amounts of essential dietary amino acids which are rare in the plants eaten. So lemmings and other grazing mammals, though they may eat voraciously, slowly starve. An enlarged pancreas is an associated symptom.

The Norwegian researchers found that levels of the inhibitor in these plants rose considerably in known years of peak lemming numbers and remained high during the following year of their decline, falling back to a very low level in the next year when lemming numbers again returned to a population low.

Lemmings taken from a declining population had pancreases nearly three times the normal size. They also had retarded growth, 25 per cent of them dying within 10 days of capture despite consuming more than 10 times their own body weight in food each day.

It is probable that the high trypsin inhibitor levels in peak and early decline lemming years can also explain the rodents' delayed sexual maturation, compressed breeding season and high dispersal rates at such times.

To cause the cyclical changes in lemming numbers, there has to be a link between high rodent numbers and the quantity of the inhibitor produced. There is, and it is related to grazing densities.

Cotton grasses and sedges grazed very occasionally produced inhibitors for only a short period of time, peaking around 30 hours after each bit of damage. But, if lemming numbers are high, and the plants are frequently grazed, more inhibitor is produced and it stays at high concentrations for as long as the grazing continues.

So low densities of lemmings can eat, digest their proteins and reproduce to their hearts' content. But once their populations start to climb, as inevitably they do, their food plants take defensive action. Mass starvation, and a frenzied search for pastures new, is the result.

If it is any consolation, the roller-coaster of lemming control at least guarantees that there will always be plants for them to live on. But, you might well ask, isn't it a wonder that evolution hasn't seen to it that the plant and the plant-eater reach a more stable accommodation? Of course not, otherwise what would the wildlife film-makers do with themselves?

(Photograph omitted)